Type 2 diabetes is a major public health problem worldwide. Progressive chronic kidney disease (CKD) in type 2 diabetes is associated with significant morbidity and mortality. There are few approaches available for the early detection of CKD in diabetes. Our primary goal of this proposal is to develop novel urine biomarkers to better predict progressive CKD in diabetics. We propose that there are better markers to predict CKD in diabetics than albuminuria. These include biomarkers that reflect tubulointerstitial injury, inflammation and fibrosis, and oxidative stress. We have identified 11 promising serum and urinary biomarkers that reflect these different pathways of injury in diabetic kidney disease. The ACCORD study demonstrated that intensive glycemic control vs. standard glycemic control over a period of 3.5 years reduced the incidence of albuminuria but did not reduce the incidence of incident CKD nor end-stage renal disease (2.1% vs. 2.2%) between the two glycemia-treatment arms. These findings require explanations and only underscore the necessity to predict incident and progressive CKD in type 2 diabetes. The ultimate goal is to identify better surrogates or therapeutic targets for kidney disease in diabetics. Thus, our specific aims for this proposal are the following: Specific Aim 1: To identify novel biomarkers predictive of incident and progressive diabetic kidney disease. Hypothesis 1a: Novel biomarkers of tubulointerstitial kidney injury and other pathways of kidney disease progression measured on baseline samples will predict incident diabetic kidney disease in ACCORD. Hypothesis 1b: Changes in biomarkers from baseline to will predict those with fast progression of diabetic kidney disease (ESRD) Hypothesis 1c: A comprehensive model incorporating novel biomarkers of tubulointerstitial kidney damage will predict incident and progressive DKD better than demographic, clinical and laboratory variables alone Specific Aim 2: Using the panel of the top 5 biomarkers developed under Specific Aim #1, to understand the pathways by which intensive glycemic control reduced microvascular outcomes (albuminuria and retinopathy) but did not reduce the definitive events of incident CKD or ESRD in the ACCORD trial. Hypothesis 2a: Intensive glycemic control will have a beneficial effect on the serum and urine biomarkers that are most predictive of progressive diabetic kidney disease at 24 months, potentially indicating that the duration of therapy was not sufficiently long to witness differences in CKD and ESRD between the two groups.